OUP user menu

Chloride channel myotonia: exon 8 hot-spot for dominant-negative interactions

D. Fialho, S. Schorge, U. Pucovska, N. P. Davies, R. Labrum, A. Haworth, E. Stanley, R. Sud, W. Wakeling, M. B. Davis, D. M. Kullmann, M. G. Hanna
DOI: http://dx.doi.org/10.1093/brain/awm248 3265-3274 First published online: 11 October 2007

Summary

Myotonia congenita (MC) is the commonest genetic skeletal muscle ion channelopathy. It is caused by mutations in CLCN1 on chromosome 7q35, which alter the function of the major skeletal muscle voltage-gated chloride channel. Dominant and recessive forms of the disease exist. We have undertaken a clinical, genetic and molecular expression study based upon a large cohort of over 300 UK patients. In an initial cohort of 22 families, we sequenced the DNA of the entire coding region of CLCN1 and identified 11 novel and 11 known mutations allowing us to undertake a detailed genotype–phenotype correlation study. Generalized muscle hypertrophy, transient weakness and depressed tendon reflexes occurred more frequently in recessive than dominant MC. Mild cold exacerbation and significant muscle pain were equally common features in dominant and recessive cases. Dominant MC occurred in eight families. We noted that four newly identified dominant mutations clustered in exon 8, which codes for a highly conserved region of predicted interaction between the CLC-1 monomers. Expressed in Xenopus oocytes these mutations showed clear evidence of a dominant-negative effect. Based upon the analysis of mutations in this initial cohort as well as a review of published CLCN1 mutations, we devised an exon hierarchy analysis strategy for genetic screening. We applied this strategy to a second cohort of 303 UK cases with a suspected diagnosis of MC. In 23 individuals, we found two mutations and in 86 individuals we identified a single mutation. Interestingly, 40 of the cases with a single mutation had dominant exon 8 mutations. In total 48 individuals (from 34 families) in cohort 1 and 2 were found to harbour dominant mutations (37% of mutation positive individuals, 30% of mutation positive families). In total, we have identified 23 new disease causing mutations in MC, confirming the high degree of genetic heterogeneity associated with this disease. The DNA-based strategy we have devised achieved a genetic diagnosis in 36% of individuals referred to our centre. Based on these results, we propose that exon 8 of CLCN1 is a hot-spot for dominant mutations. Our molecular expression studies of the new exon 8 mutations indicate that this region of the chloride channel has an important role in dominant negative interactions between the two chloride channel monomers. Accurate genetic counselling in MC should be based not only upon clinical features and the inheritance pattern but also on molecular genetic analysis and ideally functional expression data.

  • myotonia congenita
  • channelopathy
  • chloride channel
  • CLCN1

Introduction

Myotonia congenita (MC) is the commonest skeletal muscle ion channelopathy. An autosomal dominant and an autosomal recessive form of MC are described (Becker, 1977). Both forms are characterized by muscle stiffness that exhibits the ‘warm up’ phenomenon and a variable degree of muscle hypertrophy. Electromyography (EMG) usually reveals frequent myotonic discharges, which are helpful in confirming the myotonic disorder, but do not distinguish between the recessive and dominant variants. Studies of an animal model of dominant MC (the myotonic goat) revealed reduced muscle cell membrane chloride conductance as the basis of the disorder. This suggested the skeletal muscle voltage gated chloride channel as a strong candidate gene (Lipicky and Bryant, 1966). Subsequently, mutations in the skeletal muscle voltage-gated chloride channel gene CLCN1 were identified to be the cause of MC in humans (Koch et al., 1992; George et al., 1993). To date mutations have been reported widely distributed across the 23 exons of CLCN1 (Supplementary Table 1). A clear phenotype–genotype correlation has not emerged.

The chloride channel CLC-1 exists as a homodimer with each individual subunit forming a gated pore. The channel has two main gating modes referred to as the fast gate, which can open and close the two pores independently, and a slow gating mechanism or ‘common gate’ which causes deactivation of both pores simultaneously (Saviane et al., 1999). Electrophysiological studies have revealed that recessive mutations usually exert their effect by loss of function of the mutated subunit, while the mutant subunit in dominant disease has an adverse effect on the function of the co-expressed wild-type (WT) subunit, i.e. a dominant-negative effect (Pusch et al., 1995).

In this study, we had the opportunity to study the clinical and genetic features of MC in two large cohorts collectively comprising over 300 UK patients. We undertook a detailed genotype–phenotype correlation study in cohort 1. In total, we identified 23 new disease causing mutations, increasing the number of known mutations in this gene by almost 30%. The analysis of our data obtained from a combination of genetic and molecular expression studies allowed us to identify part of the H–I linker and I helix region of the chloride channel, encoded by exon 8, as being of particular importance in causing the dominant form of the disease.

Methods

Patients

In collaboration with the British Neurological Surveillance Unit (BNSU), we identified an initial cohort of 22 families from the UK with a strong clinical suspicion of MC. Inclusion criteria included a history of muscle stiffness, myotonia with ‘warm-up’ on examination and EMG detectable myotonia. Individuals were excluded if they presented with profoundly cold exacerbated, potassium-sensitive or paradoxical myotonia, features suggestive of periodic paralysis, current treatment with pro-myotonic medication (statins, fibrates or β-blockers), or if they had systemic features suggestive of myotonic dystrophy. All individuals underwent detailed clinical evaluation by one of the authors, and neurophysiological testing. Subjects were subdivided on the basis of the inheritance pattern into autosomal dominant, autosomal recessive or sporadic MC. We routinely measured creatine kinase activity (CK) and expansion in the DMPK gene associated with myotonic dystrophy type 1 was excluded prior to inclusion in this study.

Cohort 1 molecular genetic analysis

Informed consent was obtained for DNA analysis from each individual. DNA was extracted from blood using standard methods. All 23 exons of CLCN1 in each proband were analysed by bidirectional direct DNA sequencing using a Big Dye Terminator sequencing kit [Applied Biosystems (ABI)] and a 377, 3100 or 3730 automated DNA sequencer (ABI) (primer sequences available on request). DNA sequences were analysed using Sequencing Analysis and Autoassembler software (ABI).

For every candidate mutation identified, we applied PCR-based methods to screen 160 control chromosomes and DNA from available family members. This included single-strand conformational polymorphism (SSCP) analysis, where mutations were clearly detected by this method. In other cases mutations creating or destroying a restriction site were screened using restriction fragment length polymorphism (RFLP) analysis. To determine the effect of the splice site mutation c.1471 + 1G>A on mRNA, RT-PCR was performed on the sample from the proband in Family Q. Total RNA was extracted from blood using Trizol (Life Technologies, UK). Reverse transcription was performed using Superscript one-step RT-PCR system (Gibco, UK). The cDNA products were sequenced directly using the PCR primers and internal exonic primers (details available on request). All patients (cohort 1 and 2) were screened for the quadruplet nucleotide repeat expansion in intron 1 of the ZNF9 gene associated with myotonic dystrophy type 2 (DM2) using a fluorescent PCR based method. In the initial step, primers flanking the repeat sequence were used for amplification and PCR products were processed with an automated DNA analyser 3730 (ABI). Demonstration of two normal sized alleles excluded DM2. A second quadruplet primed PCR was carried out in samples where only a single allele was detected in the first PCR. Primers and PCR conditions are available from the authors on request.

Cohort 2 molecular genetic analysis

The second cohort of cases comprised 303 individuals referred to our centre with a clinical suspicion of MC. Based on the results of the analysis of cohort 1 and on the analysis of published mutations, we determined that the four commonest mutation-harbouring regions were exons 8, 11, 13 and the first half of exon 23 (exon 23a). Therefore, we analysed these four regions in cohort 2 and in an additional 160 control chromosomes by bidirectional direct DNA sequencing.

Site-directed mutagenesis and cRNA synthesis

Point mutations were introduced into the pTLN-hCLC-1 vector containing the full length WT CLC-1 cDNA by PCR based methods. Two overlapping fragments were amplified with primers containing the desired nucleotide change together with a second nucleotide change creating a new restriction site. Fragments were annealed, further amplified, digested by appropriate restriction endonucleases and ligated into the cDNA. Fragments were sequenced to confirm presence of the mutation. After transformation of TOP10 cells (Invitrogen) mutant clones were identified using the introduced restriction sites and mutations were confirmed by direct sequencing. Complementary RNA was obtained in-vitro with the mMessage mMachine kit (Ambion). Integrity and concentration of the product was checked by ethidium bromide stained gel electrophoresis and spectrophotometry.

Oocyte preparation and cell electrophysiology

Female Xenopus laevis frogs were sacrificed in accordance with the Animals (Scientific Procedures) Act 1986. Stage V or VI oocytes were isolated, stored and analysed in ND96 media [in mM: NaCl 96, KCl 2, MgCl2 1, HEPES 5, CaCl2 1.8, adjusted to pH 7.4 with NaOH, sterilized by filtration and storage solution containing 50 mg/l Gentamicin (Sigma)]. The follicular cell layer was removed enzymatically [2 mg/ml Collagenase Type XI (Sigma-Aldrich) in Ca2+ -free ND96 for 1 h]. Stage V and VI oocytes were selected and injected with 41.6 nl solution containing either water, WT RNA (2.5 or 1.25 ng), mutant RNA (2.5 ng) or both WT and mutant RNA (1.25 ng of each). Oocytes were incubated for 60 h at 10–15°C and then kept at 4°C. Standard whole cell two-electrode voltage-clamp recordings were performed at room temperature (20–22°C) 3–5 days after RNA injection (GeneClamp 500B and DigiData 1200 Interface, Axon Instruments). Recording electrodes were filled with 3 M KCL and had a tip resistance between 2–5 MΩ. Oocytes were kept at −35 mV holding potential. After an initial step to 60 mV tail currents were monitored at −80 mV following variable voltage steps between +80 and −160 mV.

Data analysis

Data acquisition, analysis and fitting was done with pCLAMP 9 (Axon Instruments) and Origin 6 (Microcal). All functional expression data are shown as mean ± standard error. To obtain comparable information on both current amplitudes as well as open probability, tail currents are presented as non-normalized currents from experiments performed on the same day. To ensure reproducibility experiments were carried out on at least two different batches of oocytes. Non-specific leak currents obtained from water-injected oocytes were averaged on the same day and subtracted from currents obtained from CLC-1 injected oocytes. Tail currents at −80 mV were fitted with a Boltzmann distribution: Embedded Image where, Imax is the (extrapolated) current at maximal activation, Io is a constant offset, V1/2 is the voltage at which half of the channels are open and dV represents the slope factor.

Statistics

Statistical analysis for clinical and electrophysiological data was done using Student's t-test or Fisher's exact test; P < 0.05 was considered a significant difference.

Results

Clinical features cohort 1

Of the 22 families, eight (36%) had autosomal dominant pedigrees, seven recessive pedigrees and seven were sporadic cases with no family history. The clinical and laboratory data on the probands in each family are summarized in Table 1.

View this table:
Table 1

Clinical data of cohort 1

PedSexMutationInheri- tanceOnset (years)Site of onsetMuscle painHyper-trophyCK (IU/l)Cold effectOther features
AFF167L (het)D3LegsNoNoN/AWorseWorse in pregnancy
BFA313V (het)D1LegsNoGen380No effectWorse in pregnancy
CFA313T (het)D4LegsYesNoN/ANo effectWorse in pregnancy
DMF297S (het)D5LegsNoGen1200Worse
EMF297S (het)D8Face, arms, legsYesNoNNo effect
FMW303R (het)D16LegsYesNo265Worse
GMW303R (het)D15LegsYesGenN/AWorse
HMF306L (het)D5LegsNoNo393No effect
IMG276D + 1437–1450delR5LegsYesGen640No effectSevere myotonia, dysphagia, dysarthria, startle weakness
JFR377X (het)R6LegsYesGen715No effectSevere, proximal weakness
KME67X (hom)R3LegsNoGen224No effect
LMC242X (hom)R4LegsNoGen300Worse
MFF413C (het)R5LegsNoGenN/AWorseDysphagia
NF1437–1450del (hom)R6LegsYesGenN/AWorse
OMA566T (hom)R16LegsNoGenN/ANo effectProximal weakness
PMF307S + F413CS4LegsNoGenN/ANo effectSevere, falls
QF1471 + 1G>A (het)S1.5LegsNoCalf344No effect
RFF307S + 1437–1450delS1LegsYesGen276WorseSJ-like syndrome, myopathic EMG
SMG285E + R894XS2LegsNoGen302WorseSevere, falls, proximal weakness
TM1183–1187del (5bp deletion) (het)S7LegsNoGen404No effect
UFR894X (het)S2LegsNoGen844Worse
VMC271R (het)S12LegsNoCalfN/AWorseDysarthria
  • Ped = pedigree; hom = homozygous; het = heterozygous; del = deletion; Gen = generalized; N/A = not available; N = normal; SJ = Schwartz-Jampel. Novel mutations highlighted in bold.

Several clinical features did not differ significantly between dominant and sporadic/recessive cases, including age of onset (mean 6.0 years), site of onset (lower limbs in all but one proband), presence of muscle pain (36% of patients), CK level (range normal—1200 IU/l) and mild worsening of myotonia in the cold (50%). Transient weakness was only reported in one patient with dominant MC but was demonstrable in approximately two-thirds of sporadic and recessive cases (P = 0.048). Other distinguishing clinical features included muscle hypertrophy (dominant versus recessive/sporadic = 38% versus 100%; P = 0.002) and diminished reflexes (dominant versus recessive/sporadic = upper limb 13% versus 55%, P = 0.048; lower limb 0% versus 20%, P > 0.05). More severe myotonia or features of persistent muscle weakness were typically seen with recessive/sporadic MC. In one case the patient (Family R) was initially considered to have Schwartz–Jampel syndrome. To illustrate the range of severity in recessive/sporadic MC more detailed descriptions of two families are available as supplementary information (Supplementary Data 1).

Molecular genetics (cohort 1)

Seven out of eight dominant families harboured mutations in exon 8. Only one mutant allele was identified in two recessive and four sporadic MC families and patients (Fig. 1).

Fig. 1

Overview of cohort 1 and 2. Numbers in brackets refer to number of families. Mutation positive individuals (asterisk) in cohort 2 classified as ‘Other’ do not necessarily have recessive MC even if a single known recessive mutation has been identified. For further details please see Results section.

Eleven novel mutations were identified (Tables 1 and 2; Supplementary Fig. 1). These included seven missense, three nonsense and one deletion. Features consistent with pathogenicity included high evolutionary conservation of the affected residues, absence in 160 control chromosomes and segregation with disease (in cases where DNA from other family members was available). In this initial cohort we did not identify pathogenic mutations in exons 1, 3, 5, 9, 12, 14 and 16–22.

View this table:
Table 2

Overview of all newly identified CLCN1 mutations

Cohort 1 (Ped)/2 (No)Nucleotide changeAA changeExonDomainModeSecond mutation identifiedClinicalFunctional
1 (K)199G>TE67X2N-termRhomSee Table 1n.d.
1 (L)726T>AC242X6FRhomSee Table 1n.d.
1 (V)811T>CC271R7GSSee Table 1n.d.
1 (I)826G>AG276D7GR1437–1450delSee Table 1n.d.
1 (D + E) and 2890T>CF297S8H–IDSee Table 1, total of seven families with 12 individuals genetically confirmedRight shift of Po, dominant negative effect
1 (F + G) and 2907T>CW303R8IDSee Table 1, total of 14 families with 19 individuals genetically confirmedDominant negative effect
1 (H) and 2916T>CF306L8IDSee Table 1, total of three families with five individuals genetically confirmedDominant negative effect
1 (B) and 2938C>TA313V8IDSee Table 1, total of three families with five individuals genetically confirmedDominant negative effect
1 (J)1129C>TR377X10JRSee Table 1n.d.
1 (T)1183–1187delfs427X11KSSee Table 1n.d.
1 (O)1696G>AA566T15QRhomSee Table 1n.d.
2 (1)866G>AS289N8HSSingle individualn.d.
2 (2)895G>CV299L8H–I?RhomUnaffected parents are distant cousins, evidence of transient weaknessn.d.
2 (3)950G>TR317L8IS1437–1450delSingle individual, parents unaffected, mother carries 1437–1450deln.d.
2 (4)959C>TA320V8IS1437–1450delSingle individualn.d.
2 (5)961G>CV321L8I–JRParents unaffected, three out of 14 siblings affected, muscle hypertrophyn.d.
2 (6)979 + 1G>AIntr. 8I–JSF413CSingle individualn.d.
2 (7)1193C>TT398I11KSR894XSingle individual, unaffected parents carry either mutationn.d.
2 (8)1205C>TA402V11KSSingle individualn.d.
2 (9)1222C>GP408A11K-LSG285E/–/–Three unrelated individualsn.d.
2 (10)1439C>AP480H13M–NSM485ESingle individualn.d.
2 (11)2596-1G>AIntr. 22C-term (CBS2)R/SR894X/–One family with two affected siblings where unaffected parents carry either R894X or 2596-1G>A; another unrelated single individual with no 2nd mutation identified yetn.d.
2 (12)2647C>AP883T23C-term?Rhom/homTwo unrelated individuals probably from. consanguineous familiesn.d.
  • Ped = pedigree; AA = amino acid; del = deletion; Intr. = intron; hom = homozygous; n.d. = not done; po = open probability.

In Family L (mild recessive phenotype), the proband and his sister were found to be homozygous for a new mutation resulting in a stop codon (C242X) that is predicted to lead to truncation of the chloride channel at helix F and arguably might have been expected to cause a more severe phenotype.

In Family R, (severe sporadic, Schwartz–Jampel-like phenotype) the proband was a compound heterozygote. She harboured a mutation previously described in both dominant and recessive MC (F307S) on one allele and a recessive mutation on the other (1437–1450del). Her unaffected daughter inherited the recessive allele.

It was notable that the F167L mutation, which has previously been reported in recessive MC (George et al., 1994; Meyer-Kleine et al., 1995), appeared to cause a dominantly inherited disorder in one of our kindreds.

RT-PCR and cDNA analysis confirmed that the splice donor site mutation detected in Family Q caused skipping of exon 13 as predicted (data not shown).

DNA-based diagnostic strategy (cohort 2)

Based on the findings from this initial cohort of 22 families and review of published mutations, we designed a strategy to streamline DNA-based diagnosis in suspected cases of MC. This consisted of screening exons 8, 11, 13 and 23a. In addition, we tested for the expansion associated with myotonic dystrophy type 2. In the 303 samples analysed, we identified CLCN1 mutations in 109 individuals (36%) (Fig. 1). Overall, 12 additional novel mutations were detected (10 missense, two splice site), which are highly likely to be pathogenic given the type of amino acid (AA) change, degree of conservation and absence in controls (Table 2; Supplementary Fig. 2). Three patients tested positive for myotonic dystrophy type 2. This included a sibling pair with no CLCN1 mutation identified and a proband that also carried the CLCN1 R894X mutation. Detailed clinical evaluation of the sibling pair by one of us (M.H.) confirmed a pure myotonic phenotype without muscle weakness, muscle pain or cardiac involvement. This interesting observation suggests that a phenotype indistinguishable from MC can occasionally associate with DM2.

Fig. 2

Tail currents of dominant mutations F297S, W303R, F306L and A313V (white squares) compared to WT CLC-1 (dark squares and circles) and WT-mutant co-expression (half dark circles). Continuous lines represent Boltzman fits. All data points represent mean ± SEM of n = 6–10. Tail currents were measured 12 ms after stepping from a 200 ms test pulse to −80 mV.

The following salient observations were made within the CLCN1 mutation positive group:

  • Twenty-three out of 109 subjects (from 22 families) carried two mutations (eight homozygous, 15 compound heterozygous) with G285E and R894X most commonly identified. This group included one patient homozygous for F307S, two patients compound heterozygous for F307S and a second recessive mutation (F413C and 1437-1450del, respectively), one patient homozygous for the (dominant) mutation A313T and one patient compound heterozygous for A313V and M485V.

  • Eighty-six out of 109 mutation-positive subjects carried a single mutation.

  • Forty of these patients (26 additional kindreds to cohort 1) harboured clearly dominant mutations (all in exon 8). The majority were the previously undescribed mutations we had identified in cohort 1 (F297S × 4, W303R × 17, F306L × 10, A313V × 4)

  • Four unrelated patients carried the F307S mutation. One of these had a clear dominant family history, one patient had an affected aunt, one patient had no family history and one was considered to be recessive.

  • Six unrelated patients harboured five different mutations not previously reported (four missense, one splice site).

  • Thirty-six patients (from 33 kindreds) harboured previously described recessive mutations including one patient who tested positive for myotonic dystrophy type 2 (G285E × 15, 1437–1450del × 5, R894X × 5, F413C × 4, V327I × 3, M485V × 2, 1471 + 1 g>a × 2).

Overall, in cohort 2 the results of inheritance pattern analysis, molecular genetic analysis and functional expression data (see also subsequently) were in keeping with a diagnosis of dominant MC in 40 patients. For the remaining patients further testing (including screening of the entire CLCN1 gene and testing of family members) will be required to allow definitive classification into recessive or dominant MC. In particular, F307S is listed separately as it can act in either an autosomal dominant or an autosomal recessive manner.

Functional expression

We studied the functional consequences of the four novel suspected dominant mutations F297S, W303R, F306L and A313V in the Xenopus oocyte expression system. WT CLC-1 channels have their maximum open probability at positive voltages and partially deactivate at negative membrane potentials. Three of the four mutant channels, W303R, F306L and A313V expressed on their own generated either profoundly reduced or no currents at all voltages (Fig. 2). Co-expressing mutant W303R or A313V and WT channels resulted in larger currents than mutant channels alone. However, these mutant subunits exhibited a strong dominant negative effect on WT channels at all potentials. F306L mutant channels co-expressed with WT showed a loss of the sigmoidal shaped relationship between voltage and tail current as well as a dominant negative effect on current amplitudes at most potentials. In contrast, F297S mutant channels had a steeper current–voltage (IV) relationship, resulting in larger currents than WT channels at strongly depolarized potentials, but significantly reduced currents at all membrane potentials in the physiological range. This shift meant that F297S had a V1/2 of −19.7 ± 0.9 mV compared to −71.1 ± 2.5 mV for WT (P < 0.0001, n = 9). Co-expression of WT and F297S showed similar conductances to expression of F297S channels alone for most of the voltages tested except for very positive voltages.

Discussion

Clinical perspective

To date, this is the largest clinical and genetic study of MC. It suggests an unexpected high proportion of dominant MC (37% of cases, 30% of families within the mutation positive group) in the two cohorts we studied. We recognize that we cannot completely exclude some form of bias introduced by ascertainment or by the limited CLCN1 screening in cohort 2. However, all studies of this kind are potentially susceptible to bias by the nature of referral patterns and prioritization of patients for genetic diagnosis. Previous estimates of the relative prevalence of dominant and recessive forms of MC indicated a lower proportion of dominant disease (Lehmann-Horn and Jurkat-Rott, 1999). In the three largest studies describing more than 15 genetically confirmed chloride channel MC patients, clear dominant MC was reported in less than 10% of mutation-positive German and Scandinavian families [summary in the discussion of Meyer-Kleine et al. (1995), Papponen et al. (1999) and Sun et al. (2001)]. If all published genetically proven MC cases are combined, the numbers suggest that ∼16% of families may harbour dominant CLCN1 mutations (for references, see Supplementary Table 1). In comparison to these previous genetic studies, our observations indicate a relatively high proportion of dominant MC. Our findings are not dissimilar to those in the original meticulous study by Becker (1977). Becker characterized the clinical features, inheritance patterns and prevalence of MC in West Berlin and West Germany, prior to the identification of the CLCN1 gene. He identified 142 families with clinical evidence for MC, of which 27 (19%, with 126 affected individuals) were dominant, 104 (73%, with 148 affected individuals) were classified as recessive and the remainder unclassifiable according to inheritance (but thought to be probably belonging to the recessive group). Becker's seminal study predates genetic testing and it is possible that inadvertent inclusion of other dominant myotonic disorders (e.g. DM2 and sodium channel disease) may have lead to an overestimation of dominant MC. Our critical review of the detailed clinical features provided by Becker suggests to us that it is most likely that the majority of his 126 dominant cases had chloride channel MC. Further study is required to fully answer the question of relative prevalence of dominant versus recessive MC and this needs a comprehensive prevalence study in a defined population.

Our study also serves to highlight the similarities in presentation of dominant and recessive MC, although there are some prominent differences. In contrast to previous studies, often without genetic confirmation (Becker, 1977), we did not find that dominant MC presents earlier than recessive MC. Lower limb stiffness was the most common presenting complaint in both dominant and recessive/sporadic MC. One third of subjects (36%) described significant muscle pain most commonly affecting the lower limbs. There were however no other specific features, with symptoms varying from pain on initiation of movement to pain at rest following exertion. Mild to moderate worsening of muscle stiffness and muscle pain with cold was reported in 50% of the patients we studied. Modest elevations of CK were common in both dominant and recessive MC, but the highest CK was 1200 IU/l, lower than in some previous studies that have demonstrated values over 2000 IU/l (Nagamitsu et al., 2000).

More severe myotonia, including bulbar dysfunction and proximal muscle weakness, was however found more commonly in recessive MC in the present study, and one patient presented with a Schwartz–Jampel-like syndrome. Generalized muscle hypertrophy was more common in recessive MC (100%) but also occurred in dominant cases (38%). Transient weakness (i.e. sudden loss of muscle power when initiating movement after rest) was strongly associated with recessive MC, present in approximately two thirds of cases, while only one patient with dominant MC had this feature. Depressed upper limb reflexes were also found in approximately two thirds of recessive cases, and only present in one patient with dominant MC.

Treatment efficacy was not formally assessed but our personal clinical observations suggest that mexiletine is often effective in MC.The rationale for using other therapies is anecdotal (Davies and Hanna, 2001). Natural history studies and treatment trials for MC are required to improve our understanding and ability to manage this condition. These are now in progress through our Consortium for Clinical Investigation of Neurological Channelopathies (CINCH) (http://rarediseasesnetwork.epi.usf.edu/cinch/index.htm).

Molecular genetic perspective

Genetic analysis in MC is problematic for a number of reasons, not least the fact that many families have ‘private’ mutations. In order to maximize mutation identification, we elected to analyse all 23 exons of CLCN1 by direct DNA sequencing in cohort 1, since our initial experience with SSCP analysis showed that direct DNA sequencing increases the mutation detection rate by ∼10%. Although mutations causing MC can occur throughout the coding region of CLCN1 we found a clear hot-spot for mutations causing dominant MC in exon 8 encoding part of the H–I interlinker and I helix of the channel. We also identified one individual with dominant inheritance pattern in their family (Family A), who harboured the F167L mutation, which has previously been reported as a heterozygous change on two occasions with presumed recessive inheritance with no second mutation identified in either proband (George et al., 1994; Meyer-Kleine et al., 1995). Functional expression has shown a small shift of the voltage dependence of the CLC-1 open probability but a dominant negative effect was not examined (Zhang et al., 2000). It is possible that F167L may result in either dominant or recessive MC and further expression studies could be helpful.

It is reported that the same mutation can cause dominant or recessive MC [for review see Colding-Jorgensen, (2005)]. Possible explanations include variable expressivity and/or penetrance and the majority of mutations demonstrate a mild dominant negative effect in co-expression studies (Kubisch et al., 1998; Plassart-Schiess et al., 1998). This point is highlighted by the A313T mutation we detected in Family C. The same mutation was previously discovered in one dominant MC family and in one apparently recessive MC kindred where the unaffected mother of the proband carried the mutation. A313T had a dominant negative effect in functional expression studies (Plassart-Schiess et al., 1998). In our second cohort we confirmed the A313T mutation mostly in other patients with dominant MC but also found it as a homozygous change in a proband from a consanguineous family with no other known affected family members. At the same codon we identified a new mutation causing a different AA substitution (A313V, Family B cohort 1) with moderately severe myotonia. Variable penetrance was not a feature in this family. Interestingly, we identified one proband with dominantly inherited MC who carried the recessive CLCN1 mutation M485V together with the A313V mutation. He was the most severely affected individual from this pedigree. Unfortunately, further samples from other family members were not available for genetic analysis.

Only one mutant allele could be identified in 6/14 (43%) of the recessive and sporadic MC families in cohort 1. There are a number of possible explanations for this. The usual reason given is that the second mutation has not been detected as it lies deep within an Intron or possibly in the promoter region of CLCN1. Firm conclusions regarding the inheritance of R377X, C271R and 1183-1187del (5 bp deletion) requires functional expression data and the study of other kindreds harbouring the same mutations. Purely on the grounds of their presence in either a compound heterozygous or homozygous state and the clinical inheritance pattern, the new mutations E67X, G276D, C242X (see subsequently) and A566T are likely to be recessive.

Three compound heterozygote combinations were detected in association with a severe MC phenotype (F307S/1437–1450del; G285E/R894X; F307S/F413C). The combination of mutations is interesting and particularly important for genetic counselling. The F307S mutation has been described in two unrelated patients. One of them was a compound heterozygote with a small deletion in the C-terminus of the channel on the second allele (Colding-Jorgensen et al., 2003). The parents of this patient were said to be unaffected. The other patient had a dominant family history and functional expression showed a dominant negative effect in co-expression studies (Kubisch et al., 1998). Our own data on this mutation from cohort 2 supports the evidence that this mutation can act dominantly in some and recessively in other pedigrees. Similarly, while G285E is known to behave in an autosomal recessive manner, R894X has also been described in dominant MC (Colding-Jorgensen et al., 2003) and has a mild dominant negative effect when expressed (Meyer-Kleine et al., 1995). Each of these three probands therefore has a mixture of a recessive and a (partially) dominant mutation.

Functional expression and structure–function implications

Electrophysiological expression studies are an invaluable tool to establish pathogenicity of ion channel mutations identified in patients with neurological diseases. For MC, in particular, it has been useful to resolve patterns of inheritance. As confirmed by X-ray crystallographic studies of bacterial CLC proteins (Dutzler et al., 2002) CLC channels exists as a homodimer with each individual subunit forming a gated pore. Although these prokaryotic CLC proteins are transporters and not channels, their structure and behaviour is thought to be closely related to that of mammalian CLC channels (Miller, 2006).

Recessive mutations often result in a complete loss of the protein due to nonsense-mediated decay, impaired transport to the membrane or inability to form dimers (typically early nonsense, frame-shifting insertion/deletions or splice-site mutations). If functional subunits are formed recessive mutations do not interfere with the function of the co-assembled WT subunit (typically missense mutations). In dominant MC, the mutant subunit has an adverse effect on the function of the co-expressed WT subunit, i.e. a dominant negative effect. The four newly identified dominant mutations in this study all showed a significant dominant negative effect in the co-expression studies in keeping with our clinical observation. While CLCN1 mutations are scattered across the whole of the channel it has been reported that dominant mutations may be more common at the dimer interface of the channel consisting of helices H, I, P and Q (Duffield et al., 2003) (Fig. 3). Interestingly, three out of the four dominant mutations (W303R, F306L and A313V) are located within the I helix and these showed little if any current when expressed on their own. In contrast, the F297S mutation located within the H–I linker region revealed more physiological current amplitudes but with a significant right-shift of the open probability curve. All mutations are located three to four AA (or a multiple of) apart from each other indicating that the affected AA are orientated towards the same side of an α-helix.

Fig. 3

Structure of a CLC subunit according to Dutzler et al. (2002). Top—2-dimensional CLC subunit with location of novel mutations indicated. Circles represent mutations from cohort 1 with dominant mutations underlined. Squares represent mutations from cohort 2, for numbers see Table 2. Bottom—3-dimensional picture of a single CLC subunit as seen from the dimer interface. The helices H, I, P and Q forming the channel interface are represented in white. The approximate location of the dominant mutations F297S, W303R, F306L and A313V are indicated as white circles. The black circle represents a chloride ion within the pore.

Conclusion

This UK study has highlighted important clinical and molecular genetic features of chloride channel MC. We observed recurrent novel dominant chloride channel mutations. Overall this study has resulted in the addition of 23 new disease-causing CLCN1 mutations to the list of pathogenic chloride channel mutations underlining the marked genetic heterogeneity of chloride channel myotonia. Although there are clearly many important functional regions in this channel our data highlights the importance of the region encoded in exon 8 judging from the high density of mutations in this region and their dominant inheritance pattern. Using the data collected in this study, in combination with published data, we have been able to develop a rational, DNA-based diagnostic service for MC in the UK.

Several of the cases presented in this article illustrate the possible pitfalls in genetic counselling of MC patients. Ideally, accurate genetic counselling in MC should take into account segregation patterns in other kindreds harbouring the same mutations and where possible functional expression data. We hope that the data presented here will be of value in aiding genetic diagnosis and genetic counselling in patients with MC.

Supplementary material

Supplementary material is available at Brain online.

Acknowledgements

We are grateful to Prof. T. J. Jentsch of the Centre for Molecular Neurobiology (Hamburg, Germany) for providing the human CLC-1 clone. This work was undertaken at UCLH/UCL who received a proportion of funding from the Department of Health's NIHR Biomedical Research Centres funding scheme. We would like to thank the referring physicians and the BNSU. This work was supported by the Wellcome Trust UK, the MRC, the Worshipful Company of Pewterers and CINCH (NIH Grant No. 1 U54 RR198442-01). Further information about the UK diagnostic service for skeletal muscle channelopathies supported by the the UK National Commissioning Group-NHS is available from Prof. M. G. Hanna mhanna{at}ion.ucl.ac.uk

Footnotes

  • Abbreviations:
    Abbreviations:
    AA
    amino acid
    ABI
    Applied Biosystems
    CK
    creatine kinase
    del
    deletion
    DM2
    myotonic dystrophy type 2
    EMG
    electromyography
    het
    heterozygous
    hom
    homozygous
    Intr.
    intron
    MC
    Myotonia congenita
    Ped
    pedigree
    po
    open probability
    SSCP
    single-strand conformational polymorphism
    WT
    wild-type

References

View Abstract